THE MEASUREMENT of cardiac output (CO) is a parameter

Transcription

1 Comparison of Esophageal Doppler, Pulse Contour Analysis, and Real-Time Pulmonary Artery Thermodilution for the Continuous Measurement of Cardiac Output Berthold Bein, MD,* Frank Worthmann, MD,* Peter H. Tonner, MD,* Andrea Paris, MD,* Markus Steinfath, MD,* Jürgen Hedderich, PhD, and Jens Scholz, MD* Objective: Continuous measurement of cardiac output (CCO) is of great importance in the critically ill. However, pulmonary artery thermodilution has been questioned for possible complications associated with right heart catheterization. Furthermore, measurements are delayed in the continuous mode during rapid hemodynamic changes. A new pulmonary artery catheter CCO device (Aortech, Bellshill, Scotland) enabling real-time update of cardiac output was compared with 2 different, less-invasive methods of CCO determination, esophageal Doppler and pulse contour analysis. Design: Prospective, observational study. Setting: University hospital, single institution. Participants: Patients scheduled for elective coronary artery bypass grafting (CABG). Interventions: None. THE MEASUREMENT of cardiac output (CO) is a parameter often used to assess the hemodynamic status and efficacy of therapy in critically ill patients. 1 In past decades, intermittent bolus thermodilution cardiac output (ICO) with ice cold saline via a pulmonary artery catheter (PAC) was the gold standard for the calculation of cardiac output according to the Fick principle in the clinical setting. 2 Over the years, interest in continuous monitoring of cardiac output (CCO) increased because the assessment of changes in the hemodynamic status of patients over time can facilitate adequate therapy. As a result, among other devices, PACs with integrated heating filaments were developed. Intermittently, this filament heats the blood up to 4 C over baseline temperature and an attached computer calculates the resulting cardiac output. 3 However, the heating filament-based CCO measurements show a lack of agreement with ICO during rapid hemodynamic changes because of the time constant of the calculation algorithm. 3,4 Recently, the value of PACs has been questioned. In a large study the PAC was found to increase mortality, hospital stay, and costs. 5 Ramsey et al 6 also found that PAC measurements had no impact on clinical decision making. Therefore, alternative measurement methods have been developed that are less invasive and/or allow the real-time calculation of CO. A new PAC with an alternative calculation principle was compared with pulse contour analysis and ultrasound-based measurements for the continuous measurement of CO. In previous studies, good-to-excellent agreement of the methods under investigation was shown with the gold standard of pulmonary arterial bolus thermodilution Therefore, the authors did not use pulmonary arterial thermodilution as the reference for these measurements, but instead compared the continuous methods with each other in the setting of a cardiac surgical unit. Measurements and Main Results: CCO measurements were analyzed using a Bland-Altman plot. Bias between CCO and pulse contour cardiac output (PCCO), and Doppler-derived cardiac output (UCCO) was (mean 1 SD) L/min versus L/min, and between UCCO and PCCO L/min. Bias was not significantly different among methods, nor were comparative values before and after cardiopulmonary bypass (p > 0.05). Conclusions: Agreement between the CCO method and both less-invasive measurements was clinically acceptable. There were no adverse events associated with the use of either device Elsevier Inc. All rights reserved. KEY WORDS: continuous cardiac output, esophageal Doppler, pulmonary artery catheter, pulse contour analysis METHODS After approval of the institutional review board committee and after written informed consent, 10 American Society of Anesthesiologists physical status IV patients with impaired left ventricular function (ejection fraction 50%) scheduled for elective cardiac surgery (coronary artery bypass grafting) were enrolled in the study. Patients with valvular heart disease, intracardiac shunts, or peripheral vascular disease, as well as emergency cases, were excluded. Only patients with sinus rhythm in the preoperative electrocardiogram were included. Patients received 0.1 to 0.2 mg/kg of midazolam and 2 g/kg of clonidine, orally, 30 minutes before the induction of anesthesia. After local anesthesia, a 5F introducer was inserted into the right femoral artery and a 4F thermodilution catheter (Pulsion Medical Systems, Munich, Germany) was placed and connected to the pressure transducer for continuous arterial pressure recording. Anesthesia was subsequently induced with propofol (2 mg/kg) and sufentanil (0.5 g/kg). Tracheal intubation was facilitated with rocuronium (0.6 mg/kg) and the patient ventilated with an air/oxygen mixture (F I O 2 0.5). Ventilation was adjusted to a PetCO 2 of 35 mmhg. A PAC (Aortech, Bellshill, Scotland) was inserted via an 8.5F introducer in the right internal jugular vein and advanced under continuous pressure recording into the wedge position. A monitor (Aortech, Bellshill, Scotland) was attached to the PAC and calibrated according to patient height and weight. This new system uses a thermistor in a small heating coil and measures the energy required to maintain the coil surface at 1 C differential above the blood temperature. This mechanism of action enables a beat-to-beat update of the CO for CCO measurement. The arterial catheter was connected to a monitor for pulse contour analysis of CO (PCCO) (Pulsion Systems) and the resulting signal processed for determination of hemodynamic variables (left ventricular stroke volume and derived parameters). To calibrate the system for the individual vascular impedance, pulse contour analysis was performed From the Departments of *Anaesthesiology and Intensive Care Medicine and Biostatistics, University Hospital Schleswig-Holstein Campus, Kiel, Germany. Presented in part at the Annual Meeting of the American Society of Anesthesiologists, Orlando, FL, October 12-16, Address for reprint requests to Berthold Bein, MD, University Hospital Schleswig-Holstein, Department of Anaesthesiology and Intensive Care Medicine, Schwanenweg 21, D Kiel, Germany. bein@anaesthesie.uni-kiel.de 2004 Elsevier Inc. All rights reserved /04/ $30.00/0 doi: /j.jvca Journal of Cardiothoracic and Vascular Anesthesia, Vol 18, No 2 (April), 2004: pp

2 186 BEIN ET AL Fig 1. Bland-Altman plot between CCO and PCCO. The solid line represents the mean difference (bias); the dotted line represents the 2SD limits of agreement. while simultaneously injecting 10 ml of ice cold saline in the proximal port of the PAC and registration of the resulting arterial thermodilution curve via the catheter in the femoral artery. The mean of three consecutive measurements randomly assigned to the respiratory cycle was used for calibration. The PCCO was not recalibrated during the surgical procedure. An esophageal echo probe (Hemosonic, Arrows International, Everett, MA) was inserted nasally and advanced into the esophagus approximately up to the sixth thoracic vertebra for measurement of ultrasound Doppler cardiac outputs (UCCOs). Depth of probe insertion in each patient was chosen to obtain the best signal quality, and, therefore, the position may have varied in the individual patient. Two acoustic transducers are located at the tip of the flexible esophageal probe. The echo signal was adjusted to the maximum signal height and the probe positioned until both the anterior and the posterior wall of the aorta were visible on the screen. The echo probe was readjusted if necessary (loss of the aortic wall detected by M-mode ultrasound). Before and after cardiopulmonary bypass (CPB), 6 measurements were performed. Timing of data acquisition was assigned to 12 different time points, which were defined as follows: (1) baseline after induction of anesthesia, (2) skin incision, (3) sternotomy, (4) start harvesting of the mammary graft, (5) end harvesting of the mammary graft, (6) before initiation of CPB, (7) directly after termination of CPB, (8) 15 minutes after termination of CPB, (9) start thoracic closure, (10) end of thoracic closure, (11) end of surgery, and (12) before discharge to the intensive care unit. These time points were at least 15 minutes apart and were chosen to achieve a high within-subject variability in cardiac output. CO was measured during stable hemodynamic conditions. All coronary artery bypass grafting operations were performed uniformly using a standard CPB technique (pump flow rate of 2.5 L/min/m 2 ), with mild hypothermia (rectal temperature C). Statistical analysis was performed according to the method of Bland and Altman. 7 Bias between methods was calculated as the mean difference ( SD) between CCO and PCCO, between CCO and UCCO, and between PCCO and UCCO. The limits of agreement were defined as bias 2 SD and as the range in which 95% of the differences between the methods were expected. Data points from each individual were averaged; resulting mean values were then compared for betweenmethod differences with analysis of variance for repeated measures with Bonferroni correction. Bias before and after the cardiopulmonary bypass was analyzed with paired student t test. Statistical significance was assumed at a value of p RESULTS Ten patients (aged years; 6 male, 4 female) were enrolled in the study. A total of 113 PCCO, 107 UCCO, and 113 CCO measurements were analyzed. CO measurements ranged from 1.89 to 8.6 L/min for PCCO, 1.5 to 8.2 L/min for UCCO, and 2.4 to 5.7 L/min for CCO. The Bland-Altman plot for CCO and PCCO is shown in Figure 1, for UCCO and CCO in Figure 2, and for PCCO and UCCO in Figure 3. Bias between CCO and PCCO was 0.71 L/min (precision 1 L/min), between CCO and UCCO 0.15 L/min (precision 1.09 L/min), and between UCCO and PCCO 0.58 L/min (precision 1.06 L/min). Linear regression analysis of the CCO/PCCO and CCO/UCCO Bland-Altman plot yielded a negative slope representing an overestimation of low COs and an underestimation of high COs compared with PCCO and UCCO (r and 0.27, respectively). Bias between methods showed no significant differences (p 0.05, Fig 4). Comparing values before and after CPB, bias of UCCO, PCCO, and CCO measurements did not differ significantly (p 0.05, Fig 5). There were no adverse effects related to either the PCCO/ PAC catheter or the echo probe. DISCUSSION Perioperative determination of cardiac output is of great interest in the critically ill. Since 1970, PAC thermodilution has become the clinical gold standard in the field of anesthesia and intensive care. However, right heart catheterization for CO monitoring has been questioned for various reasons. First, ICO shows remarkable variance and has proved to be no real reference method in comparison studies. 8,9 Second, until recently,

3 ESOPHAGEAL DOPPLER VERSUS PCCO VERSUS PAC 187 Fig 2. Bland-Altman plot between CCO and UCCO. The solid line represents the mean difference (bias); the dotted represents the 2SD limits of agreement. there was no real CCO measurement by PA thermodilution. The PAC with integrated heating filament was able to perform semicontinuous determinations at best. The time constant of the measurement algorithm led to a delayed time response in cases of rapid hemodynamic changes. 3,4 Unfortunately, there is no clear evidence in the literature for the need for CO determinations. Treating patients with the goal of augmenting CO has given inconsistent results with some authors reporting no improvement in survival rates and others finding reduced mortality and duration of hospital stay. 10,11 Reviewing the current literature, it seems that at least in a subset of patients, measurement of CO is still indicated for guiding adequate therapy. 12 Because invasiveness of right heart catheterization is under debate, alternative measurement methods were developed that are less invasive and/or allow the real-time calculation of CO. Impedance cardiography, partial CO 2 rebreathing, esophageal Doppler, and pulse contour analysis are newly developed or enhanced, each having specific drawbacks and advantages. Recently, a new PAC was introduced into clinical practice. Bias was found to be acceptable under clinical conditions compared with pulse contour and Doppler-derived values, ( L/min and L/min, respectively), in the present study. Esophageal Doppler-derived CO measurements have given inconsistent results in the literature. Initial studies showed significant variability between Doppler-derived and thermodilution measurements, and the technique was found to be clin- Fig 3. Bland-Altman plot between UCCO and PCCO. The solid line represents the mean difference (bias); the dotted line represents the 2SD limits of agreement.

4 188 BEIN ET AL Fig 4. mean). Bias between methods (mean standard error of the ically unacceptable because of operator dependency and the frequently necessary readjustment of the echo probe. 13,14 In this study, a newly developed echo probe (Hemosonic; Arrow International) was used, offering the advantage of determining the true aortic diameter by M-mode ultrasound, thus avoiding errors introduced by nomogram-derived calculations. Measurements performed with this device are in good agreement with PA thermodilution CO in preliminary investigations. 15 The authors confirmed good agreement of Doppler-derived measurements with both CCO and PCCO. However, readjustment of the echo probe was commonly necessary after CPB, although the same experienced investigator positioned the echo probe. The authors cannot completely exclude the influence of a learning curve on these results. Lefrant et al 16 found a remarkable training effect using the device in intensive care unit patients. The operator dependency remains a problem, particularly in the case of readjustment. The echo probe measures aortic blood flow, which is closely correlated with CO (r 0.89). 17 This derived cardiac output was used in the study calculations to facilitate between-method comparisons. There are various other parameters derived from the aortic blood flow (eg, peak ejection velocity and left ventricular ejection time), which give additional information on left ventricular performance and may help to guide optimal therapy. Pulse contour analysis has gained widespread attention recently. Arterial pulse pressure waveform analysis according to the method by Wesseling consists of measuring the area under the systolic portion of the arterial pulse wave from the end of diastole to the end of the ejection phase. 19 Numerous studies have shown good agreement with arterial and pulmonary arterial thermodilution. 20,21 Arterial cannulation is less invasive and has proven to have no severe complications during its use over a longer period of time. 18 Pulse contour enables a beat-to-beat update of the instantaneous CO. Transpulmonary thermodilution also gives important information concerning the patient s volume status and left ventricular loading conditions (eg, intrathoracic blood volume, extravascular lung water). Some authors postulated an influence of changes in systemic vascular resistance (eg, after vasopressor administration) on PCCO. 21,22 In contrast, Della Rocca et al 23 found no influence on the accuracy of PCCO even after substantial changes of SVR. The authors did not control for changes in the tone of the vascular bed because the PCCO device was not recalibrated during the study. In conclusion, by comparing a new CCO PAC and an esophageal UCCO probe with PCCO for the continuous measurement of CO in cardiac surgical patients, a clinically acceptable agreement was found between methods. However, judgment of bias and precision is subjective and not yet standardized. Critchley 19 recommended that limits of agreement between methods should not exceed 30%. Zöllner et al 20 postulated limits of agreement of 0.5 L/min between methods and rejected the interchangeability of continuous and intermittent PA thermodilution using IntelliCath (Baxter, Irvine, CA) and Opti-Q (Abbott Laboratories, Morgan Hill, CA) catheters. In a recently published study comparing PCCO and ICO, however, the same authors found limits of agreement of 2.5 L/min to be acceptable. 21 When reviewing recently published large studies comparing PCCO and ICO, bias varied between and 0.31 L/min These studies, however, compared PCCO with ICO, whereas the present study compared CCO with PCCO. Della Rocca et al 23 reported a bias of 0.03 L/min with limits of agreement of 1.78 to 1.72 L/min comparing PCCO versus CCO. With CCO, measurements show a tendency of underestimation of CO compared with ICO, possibly because of the temperature shift in the PA after CPB. 29 This may explain the negative bias between CCO and both PCCO and UCCO because PCCO and UCCO measurements are not affected by temperature. Comparison of CCO and PCCO, as well as CCO and UCCO, in the present study showed a negative slope of the Bland-Altman plot. An overestimation of low CO was also shown for the bolus thermodilution technique, probably caused Fig 5. UCCO-PCCO, CCO- PCCO, and CCO-UCCO bias ( standard error of the mean) before and after cardiopulmonary bypass. Dotted line represents an ideal bias of zero.

5 ESOPHAGEAL DOPPLER VERSUS PCCO VERSUS PAC 189 by the cold saline bolus. 31 During continuous measurement with the heating coil technique, temperature shifts in the PA commonly seen after CPB may have caused this error. This systematic deviation of the new CCO catheter has to be verified in larger study populations. Taking into account the recent discussion on the safety of PACs, several alternatives exist when CO needs to be monitored. The results of this study show that accuracy of CO measurements is clinically acceptable for all methods. Therefore, it seems unjustified to perform right-heart catheterization simply for the determination of cardiac output. The less-invasive esophageal probe and the PCCO system offer specific additional information concerning important parameters of the cardiovascular system. Hence, the choice of the equipment for CO measurement should be made according to individual patient needs. 1. Jellema WT, Wesseling KH, Groeneveld AB, et al: Continuous cardiac output in septic shock by simulating a model of the aortic input impedance: A comparison with bolus injection thermodilution. Anesthesiology 90: , Forrester JS, Ganz W, Diamond G, et al: Thermodilution cardiac output determination with a single flow-directed catheter. Am Heart J 83: , Siegel LC, Hennessy MM, Pearl RG: Delayed time response of the continuous cardiac output pulmonary artery catheter. Anesth Analg 83: , Aranda M, Mihm FG, Garrett S, et al: Continuous cardiac output catheters: Delay in in vitro response time after controlled flow changes. Anesthesiology 89: , Connors AF Jr, Speroff T, Dawson NV, et al: The effectiveness of right-heart catheterization in the initial care of critically ill patients. SUPPORT Investigators. JAMA 276: , Ramsey SD, Saint S, Sullivan SD, et al: Clinical and economic effects of pulmonary artery catheterization in nonemergent coronary artery bypass graft surgery. J Cardiothorac Vasc Anesth 14: , Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1: , Espersen K, Jensen EW, Rosenborg D, et al: Comparison of cardiac output measurement techniques: thermodilution, Doppler, CO 2 - rebreathing and the direct Fick method. Acta Anaesthesiol Scand 39: , Mackenzie JD, Haites NE, Rawles JM: Method of assessing the reproducibility of blood flow measurement: Factors influencing the performance of thermodilution cardiac output computers. Br Heart J 55:14-24, Gattinoni L, Brazzi L, Pelosi P, et al: A trial of goal-oriented hemodynamic therapy in critically ill patients. SvO 2 Collaborative Group. N Engl J Med 333: , Tuchschmidt J, Fried J, Astiz M, et al: Elevation of cardiac output and oxygen delivery improves outcome in septic shock. Chest 102: , Pinsky MR: Why measure cardiac output? Crit Care 7: , Freund PR: Transesophageal Doppler scanning versus thermodilution during general anesthesia. An initial comparison of cardiac output techniques. Am J Surg 153: , Spahn DR, Schmid ER, Tornic M, et al: Noninvasive versus invasive assessment of cardiac output after cardiac surgery: Clinical validation. J Cardiothorac Anesth 4:46-59, Odenstedt H, Aneman A, Oi Y, et al: Descending aortic blood flow and cardiac output: A clinical and experimental study of continuous oesophageal echo-doppler flowmetry. Acta Anaesthesiol Scand 45: , Lefrant JY, Bruelle P, Aya AG, et al: Training is required to improve the reliability of esophageal Doppler to measure cardiac output in critically ill patients. Intensive Care Med 24: , 1998 REFERENCES 17. Boulnois JG, Pechoux T: Noninvasive cardiac output monitoring by aortic blood flow measurement with the Dynemo J Clin Monit Comput 16: , Scheer B, Perel A, Pfeiffer UJ: Clinical review: Complications and risk factors of peripheral arterial catheters used for haemodynamic monitoring in anaesthesia and intensive care medicine. Crit Care 6: , Critchley LA: A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit 15:85-91, Zollner C, Goetz AE, Weis M, et al: Continuous cardiac output measurements do not agree with conventional bolus thermodilution cardiac output determination. Can J Anaesth 48: , Zollner C, Haller M, Weis M, et al: Beat-to-beat measurement of cardiac output by intravascular pulse contour analysis: A prospective criterion standard study in patients after cardiac surgery. J Cardiothorac Vasc Anesth 14: , Buhre W, Weyland A, Kazmaier S, et al: Comparison of cardiac output assessed by pulse-contour analysis and thermodilution in patients undergoing minimally invasive direct coronary artery bypass grafting. J Cardiothorac Vasc Anesth 13: , Della Rocca G, Costa MG, Pompei L, et al: Continuous and intermittent cardiac output measurement: Pulmonary artery catheter versus aortic transpulmonary technique. Br J Anaesth 88: , Gödje O, Höke K, Lamm P, et al: Continuous, less invasive, hemodynamic monitoring in intensive care after cardiac surgery. Thorac Cardiovasc Surg 46: , Gödje O, Thiel C, Lamm P, et al: Less invasive, continuous hemodynamic monitoring during minimally invasive coronary surgery. Ann Thorac Surg 68: , Gödje O, Friedl R, Hannekum A: Accuracy of beat-to-beat cardiac output monitoring by pulse contour analysis in hemodynamically unstable patients. Med Sci Monit 7: , Gödje O, Höke K, Goetz AE, et al: Reliability of a new algorithm for continuous cardiac output determination by pulse-contour analysis during hemodynamic instability. Crit Care Med 30:52-58, Gödje O, Höke K, Lichtwarck-Aschoff M, et al: Continuous cardiac output by femoral arterial thermodilution calibrated pulse contour analysis: Comparison with pulmonary arterial thermodilution. Crit Care Med 27: , Rauch H, Muller M, Fleischer F, et al: Pulse contour analysis versus thermodilution in cardiac surgery patients. Acta Anaesthesiol Scand 46: , Rödig G, Prasser C, Keyl C, et al: Continuous cardiac output measurement: pulse contour analysis vs thermodilution technique in cardiac surgical patients. Br J Anaesth 82: , Tournadre JP, Chassard D, Muchada R: Overestimation of low cardiac output measured by thermodilution. Br J Anaesth 79: , 1997